overWshing. Australia is not included in Rodden’s analysis and is a case in point—
the Australian states rely on central grants for half their revenue but are also
constrained by strong central controls. Thus, whether federalism is associated
with smaller or larger government depends on the mix of political and institutional
factors of particular countries.
Political scientists and policy analysts have been probing other political and
institutional factors that shape processes and outcomes in federal systems. A recent
Wnding is that political-institutional variables—the proximity of elections, the
ideology of incumbent governments, and the severity of formal rules limiting
deWcits—all have a signiWcant eVect on budgetary outcomes (Petry 2004 , 222 ).
This conclusion is based upon pooled evidence over the past couple of decades for
Wve federations. In this and other studies, Canada and Germany stand out as high
deWcit countries, while Australia, Switzerland, and the United States are low deWcit
countries.
DiVerent types of intergovernmental institutions aVect federal Wscal policy-
making in diVerent ways, as Dietmar Braun ( 2003 , 2004 ) shows using case studies
of Canada, Germany, Belgium, and Switzerland. He identiWes Canada and Ger-
many as opposite federal types—‘‘interstate’’ and ‘‘intrastate,’’ respectively—and
explains how their institutional diVerences are played out inWscal policy processes
and outcomes. Canada’s national government has extensive scope forWscal policy-
making but weak implementation because provinces are independent with their
own legislative powers. The federal government can gain leverage through provid-
ing incentives such as contributing to shared cost programs, or it can cut its
expenditure and reign in provincial spending through withdrawing from shared
programs. Whereas Canada has a competitive tax system, albeit with a shared
collection arrangement for income tax, Germany has a cooperation one (Braun
2003 , 118 ). Germany’s intrastate federalism incorporates the La ̈nder in national
Wscal policy via the Bundesrat that ensures consensus but favors the status quo, and
facilitates implementation because everyone has agreed (Braun 2004 , 25 – 8 ).
One of the main concerns with federalism, that fuelled the opposition of many
left-wing parties and commentators in the mid-twentieth century, was its conser-
vative character in favoring the status quo and making reform and innovation
diYcult. A new study by Obinger, Leibfried, and Castles ( 2005 ) shows the com-
plexity of federalism’s interaction on social policy in ‘‘new world,’’ Australia,
Canada, and the United States, and European federations, Austria, Germany, and
Switzerland. Using historical case studies, theyWnd that federalism impeded social
welfare policy early on, but after consolidation in mature systems other
cross-national diVerences explain variations among countries. The ways in which
federalism aVects policy innovation and development are multiple and complex,
variable over time, and contingent on particular institutional conWgurations,
political actors, and pressure groups, as well as broader historical and cultural
contexts. Federalism provides multiple veto-points (Tsebelis 2002 ) that can check
comparative federalism 275